agentleFS
Sign inStart free
ptrandevptrandev/claude-skills/skills/claude/SKILL.md

claude

Skill1 starsChanged 15 days agoMITView on GitHub

Calls the Claude Code CLI for an independent review, adversarial challenge, or codebase consultation. Use when the user asks Claude for a second opinion or when another skill needs an independent Claude reviewer from Codex.

35 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use compare-screenshots, implement-spec-with-codex and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • transitive-bullshitSkill

    compare-screenshots

    Compare screenshots against the intended design, distinguishing approved references from historical baselines. Use for iterative implementation matching, before/after visual review, objective image telemetry, or checking a lone capture for flat, empty, or badly framed content.

    Also in 5 projects that use this

  • dzhngSkill

    implement-spec-with-codex

    Run [implement-spec](../implement-spec/SKILL.md) with Codex doing the implementation passes while you orchestrate, integrate, and review. Use when the user asks to implement a spec with Codex, or to run implement-spec with Codex as the implementer.

    Also in 4 projects that use this

  • dzhngSkill

    close-spec

    When a spec under specs/<feature>/ is done shipping (a write-spec build, or any planned task that produced a spec), archive it to specs/done/ and rewrite it from a build-plan into a durable rationale — the why, the principles, the invariants — pointing back to the real code for the how. Use when implementation has landed and the plan no longer matches what shipped, or the user says a feature/spec is finished. Pairs with [write-spec](../write-spec/SKILL.md) (the plan this closes) and [review](../review/SKILL.md) (run before closing).

    Also in 4 projects that use this

  • wolverin0Skill

    audit-decisions

    Standardize remediation decisions, baseline-policy entries, and legal-review banners for accepted, deferred, or policy-blocked audit findings.

    Also in 4 projects that use this

  • bergsideSkill

    ant

    Structured, enterprise-focused design system emphasizing clarity, consistency, and efficiency for data-dense web applications.

    Also in 3 projects that use this

  • bergsideSkill

    artistic

    High-contrast, expressive style with creative typography and bold color choices for visually striking interfaces.

    Also in 3 projects that use this

  • +4 more from dzhng
  • +2 more from bergside

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 8 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 29 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run Claude in a fresh, non-interactive, read-only process.

    line 11
  2. Distinctive:

    Treat the text accompanying the skill invocation as the input:

    line 17
  3. Very specific:

    Use review when another skill requests an independent Claude reviewer.

    line 25
  4. Distinctive:

    Stop with the install/auth error when the command is unavailable or the CLI reports that authentication is required.

    line 30
  5. Distinctive:

    Create a temporary prompt file outside the repository.

    line 32
  6. Very specific:

    Run scripts/run-claude from that directory:

    line 35
  7. Distinctive:

    Treat timeout, empty output, or output that misses the requested contract as a failed reviewer.

    line 45
  8. Distinctive:

    Remove the temporary prompt file.

    line 47
  9. Very specific:

    When a rubric is supplied, pass its file path with --rubric and tell Claude to read it in full before reviewing.

    line 65
  10. Distinctive:

    Ask Claude to attack assumptions, boundary conditions, failure recovery, concurrency, security, and data-shape mismatches.

    line 70
  11. Distinctive:

    Ask the user's question directly.

    line 76

“Common” is measured: kept by at least 6 in 10 of the 29 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Claude

Run Claude in a fresh, non-interactive, read-only process. Do not pass this conversation, implementation rationale, or the author's identity unless the user explicitly asks for that context.

Input

Treat the text accompanying the skill invocation as the input:

InputMode
review [focus]Review the current diff.
challenge [focus]Break the current diff and its assumptions.
Any other textConsult Claude about that question.

Use review when another skill requests an independent Claude reviewer.

Run

  1. Confirm claude exists with command -v claude. Stop with the install/auth error when the command is unavailable or the CLI reports that authentication is required.
  2. Create a temporary prompt file outside the repository. Include the selected role, exact scope, evidence requirements, and output contract.
  3. Locate this skill's directory from the path used to load this SKILL.md.

  4. Run scripts/run-claude from that directory:

    <claude-skill-dir>/scripts/run-claude \
  --repo "$(git rev-parse --show-toplevel)" \
  --prompt-file <temporary-prompt-file> \
  [--rubric <rubric-file>] \
  [--model <model>]
  5. Treat timeout, empty output, or output that misses the requested contract as a failed reviewer. Report the failure instead of inventing findings.
  6. Remove the temporary prompt file.

The runner disables Claude customizations and permits only repository reads and read-only Git commands. Never weaken those controls for a review or challenge.

Review contract

Tell Claude to inspect the diff itself and verify each finding against the surrounding code. Require exactly one finding per line:

SEVERITY | file:line | one-line finding | one-line proof

Allow HIGH, MEDIUM, and LOW. Require CLEAN as the only output when no finding survives verification. Claude reviews only. It does not edit, commit, push, post, or resolve anything.

When a rubric is supplied, pass its file path with --rubric and tell Claude to read it in full before reviewing.

Challenge contract

Ask Claude to attack assumptions, boundary conditions, failure recovery, concurrency, security, and data-shape mismatches. Require the review contract above so callers can adjudicate the output mechanically.

Consult contract

Ask the user's question directly. Tell Claude to cite repository files and lines for repository claims. Keep the process read-only.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free